Hello Friends and Professors,
I have data spanning from 2010-2016 on wholesale and retail price of rice in US$. I want to use time series approach to investigate the trend in rice price(s). Also, I'd like to know whether I can average the two (i.e. wholesale price and retail price) to create a new variable as average rice price; and thereafter, use a univariate Time Series model to predict the price of rice.
Kindly note that, I have different types of rice in my data. Can I convert the short time period to monthly or quarterly?
Please help me out...
Joe
I think you can use comparative approach for both prices (wholesaler and retailer prices) during the given period.
I think you have to be careful with your analysis, particularly handing of the data and the method(s) apply. This is because, in a short series analysis, single observation could influence the overall result of the research.
I will suggest that you add other methods to your research as a comparative analysis. This will help you in having analysis with the most robust results (least prone to errors) and methods. I will suggest you add other methods from the standard/linear time series methods ( e.g AR, MA, ARMA etc ) and the nonlinear models ( e.g threshold model etc ).
Your research will become clumsy and meaningless if you add the wholesale and retail prices as an average price (New price) in your research. Just use the wholesale and/or retail prices in a univariate, bivariate or multivariate experiment.
I am not quite clear on what you want to convert. But you can use the monthly data. But short time period could be in a day, weeks, months or year.
Good luck
- Badges
- Science topic